Campaign production runs six to 12 months and six to seven figures, and most of that stretch is logistics rather than creative work. AI Central's walkthrough of Luma's Uni-1 collapses it into five steps: brief the campaign in plain language, upload references, let the model plan composition before it renders, composite multiple assets into a single scene, then localize per market without starting over. The claim is hours, not months.
Reviewed August 2026.
Where the six months actually go
The traditional path AI Central lays out is deliberately unglamorous. A campaign travels from brief to talent to shoot to post-production to localization to delivery, and every handoff sits behind a queue owned by somebody else.
AI Central puts a duration on each stage, and the shape of that list says more than the total does.
- Writing the brief, two weeks.
- Talent and locations, four weeks.
- The shoot itself, two weeks.
- Post-production, six weeks.
- Localization, eight weeks.
The shoot, the part everyone pictures when they hear the word campaign, is the shortest stage on the board. Localization is the longest, at four times the shoot. AI Central's conclusion is that most of that time is not creative at all, it is logistics, and typical budgets land in the six to seven figure range for the privilege.
What reasoning before generation actually means
The tool at the heart of AI Central's workflow is Uni-1, which Luma bills as the reasoning image model. Luma's own description of it is compact.
Uni-1 is a multimodal reasoning model that can generate pixels.
Luma extends that in the same breath, describing a model built on what the company calls Unified Intelligence.
Built on Unified Intelligence, Uni-1 understands intention, responds to direction, and thinks with you.
Strip the marketing and the mechanical claim is specific. AI Central's summary is that Uni-1 plans composition and layout before it renders anything, which is what keeps lighting, shadows and spatial accuracy consistent, and what makes the results structured and intentional rather than random images.
That ordering is the whole argument. A model that renders first and reasons never is fine for a mood board and hopeless for a campaign, because a campaign is a set, not a picture. Twelve assets that each look good on their own but disagree about where the light is coming from will not survive a brand review. Planning geometry before pixels is the difference between an image you like and an image you can ship next to eleven others.
The five steps
AI Central's process runs in five moves, and each one has a stated output rather than a vague promise.
Step one is the brief. Open Luma, select Uni-1, describe the campaign in plain language, then upload references: product shots, sketches, brand assets. AI Central's tip at this stage is to be clear about style, mood and goal, which is precisely the part most people skip.
Step two is hero visuals. Uni-1 plans composition before generating and returns polished hero images built from the brief and the references, with layout, lighting and spatial consistency held steady. The stated result is high-quality visuals ready for iteration, not finished work.
Step three is compositing. Upload multiple assets or images and Uni-1 merges them into realistic, unified scenes, holding lighting, shadows and perspective together. AI Central's stated outcome is scenes ready for campaigns in one pass.
Step four is scale. Adapt backgrounds, colorways and text per region so that one brief produces multiple outputs for different markets while brand and creative consistency hold.
Step five is export and deploy. Download the hero visuals and the variations, hand them straight to marketing, social or production teams, and launch in hours rather than months. Where a variation still is not right, AI Central points to Luma Agents, which refine and iterate while keeping the context of the original brief.
Localization is where the leverage is
Read AI Central's own timeline back and the interesting number is not the shoot. It is the eight weeks sitting at the end.
Localization is slow in traditional production because it is a restart. New market, new copy, new colorway, and often a return trip through post-production with a file that has already been flattened twice. The workflow AI Central describes attacks exactly that, localizing outputs across markets without restarting, because the model still holds the brief and the references that produced the original.
That is a structural claim, not a speed claim. It means the marginal cost of the fifteenth market falls close to the marginal cost of the second, which is the economics that make a campaign worth running in fifteen markets in the first place. Teams who only ever shipped in three markets because localization ate the budget are the ones who should be paying attention here.
What to do with this
The honest way to test the claim is to reproduce something you have already paid for, so you have a real benchmark to argue with.
- Take a campaign you shipped last year, feed the original brief and the same reference pack into Uni-1, and compare the hero output against what you actually ran.
- Write the brief for intent, not instruction. AI Central's closing direction is to describe what you want, not how to get it, which is the opposite of the prompt-craft habits carried over from earlier image tools.
- Bring real references. Product shots, sketches and brand assets are the material the model composites against, and a thin reference pack is the fastest route to generic output.
- Judge the set, not the image. Generate the full market spread before deciding, because consistency across variations is the specific thing this model claims to solve.
- Keep people on selection and approval. The workflow removes production time, it does not remove taste.
Uni-1 is free to try on Luma's site and Luma publishes a technical report alongside it, so the cost of running that test is an afternoon rather than a procurement cycle.
What is Uni-1?
Uni-1 is Luma's image model, positioned by the company as the reasoning image model and described as a multimodal reasoning model that can generate pixels. The practical difference from a standard image generator is that it plans composition and layout before rendering, which is what holds lighting, shadows and spatial accuracy consistent.
How long does a campaign take with this workflow?
AI Central puts the traditional path at six to 12 months across brief, talent and locations, shoot, post-production and localization, and puts the Uni-1 path at hours. The five steps are the brief, hero visuals, compositing, scaling across markets, then export and deploy.
Can it combine my own product photos into one scene?
Yes, that is step three. You upload multiple assets or images and Uni-1 composites them into a realistic, unified scene while holding lighting, shadows and perspective consistent, with the stated outcome being scenes ready for campaigns in one pass.
How do I adapt one campaign for different markets?
You adapt backgrounds, colorways and text per region from the same brief, so a single brief produces multiple outputs for different markets without restarting the job, and brand and creative consistency carries across them. Luma Agents handle further iteration if a variation needs another pass.
What should my brief actually say?
Plain language, with style, mood and goal stated explicitly, plus references attached. AI Central's guidance is to describe what you want rather than how to get it, and to treat clarity about mood and goal as the thing that determines output quality.